Ted Faber wrote:
Sure, but you mentioned that failure mode, and it's easy enough to
correct that mode.
I guess your point is that some people fix an issue
without looking at the bigger picture. My point was that
since there was an issue, one might as well do a better
job than a point patch. Since I think network stack
implementors normally should do a better job, I inferred
that the issue being discussed should not be there in
practice. Well, I guess you are trying to prove me wrong
that some stack implementors are not that good anyway.
Point taken (-:
tcpm mailing list